Concrete is a conglomerate of aggregate (such as gravel, sand, and/or crushed stone), water, and hydraulic cement (such as portland cement), as well as other components and/or additives. Concrete is generally fluidic when it is first made, enabling it to be poured or placed into shapes, and then later hardens, and is never again fluidic, in the general sense. Typically, moisture present in normal concrete is basic (that is, has a high pH). Alkali materials can be supplied by the cement, aggregate, additives, and even from the environment in which the hardened concrete exists (such as salts placed on concrete to melt ice).
Silica compounds are typically found in the aggregate components of concrete. Silica which is present in aggregates used to make concrete and mortars is subject to attack and dissolution by hydroxide ions present in basic solutions. Generally, the higher the pH (i.e., the more basic the solution), the faster the attack.
Different forms of silica show varying degrees of susceptibility to this dissolution. If there is sufficient alkali metal ion also present in this solution (such as sodium or potassium ions), the alkali metal ions can react with the dissolved silica and form an alkali-silica gel. Under certain conditions, the resultant alkali-silica gel can absorb water and swell. The swelling can exert pressures greater than the tensile strength of the concrete and thus cause the concrete to swell and crack. This process (hydroxide attack of silica, followed by reaction with alkali such as sodium and potassium) is referred to generally in the art as an "alkali-silica reaction" or "ASR".
ASR has caused the failure of concrete structures, although rarely. Further, ASR can weaken the ability of concrete to withstand other forms of attack. For example, concrete that is cracked due to this process can permit a greater degree of saturation and is therefore much more susceptible to damage as a result of "freeze-thaw" cycles. Similarly, cracks in the surfaces of steel reinforced concrete can compromise the ability of the concrete to keep out salts when subjected to deicers, thus allowing corrosion of the steel it was designed to protect.
ASR is a common chemical process in many concretes around the world. As an indication of its importance to the concrete industry, by 1991 over 1,450 research articles had been published on the subject. See S. Diamond, Alkali-aggregate reactions in concrete: an annotated bibliography 1939-1991, Washington, D.C.: National Research Council, Strategic Highway Research Program, SHRP-C/UWP-92-601:470 (1992).
In 1987, Congress authorized a $150 million, five-year research program to be administered by the National Research Council to study and develop improvements in highway construction materials and construction practices. This program was called the Strategic Highway Research Program (SHRP). One of the areas addressed by this program was ASR mitigation. Four recommendations resulted from the SHRP research for preventing ASR in concrete. D. Stark, et al., Eliminating or minimizing alkali-silica reactivity, Washington, D.C.:National Research Council, Strategic Highway Research Program, SHRP-C-343 (1993) (the "SHRP report").
One recommendation was the use a low alkali cement, which is defined as a cement with a sodium equivalent of 0.60% or less. The sodium equivalent of a cement is the weight percent of sodium, reported as sodium oxide, plus 0.658 times the weight percent of potassium, reported as potassium oxide. Sodium equivalent (Na.sub.2 O.sub.e) can be represented generally by the formula NaO.sub.2 +0.658.times.K.sub.2 O=Na.sub.2 O.sub.e.
While the use of a low alkali cement can have some effectiveness, it is not a guarantee of ASR prevention. Low alkali cement is not always available on a local basis, can have limited availability, and can be more expensive than high alkali cement. Further, if the raw feed for the cement production contains high levels of alkali, then the production of low alkali cement from such feed can generate much greater waste than would otherwise be the case. Generally, "fines" are a waste product of cement production and are normally recirculated into the cement kiln. However, when the raw feed has a high alkali level, the fines must be removed from the process to make a low alkali cement and constitute a waste material. These fines are generally referred in the industry as cement kiln dust.
Still further, using a low alkali cement is no guarantee of ASR control, as the cement is not the sole source of the alkalies in concrete that can contribute to the problem. Alkalies also can be supplied by pozzolans that are either admixed in or part of the blended cement. Alkalies can be supplied by the mix water, admixtures used in the concrete, the aggregate itself, including recycled concrete used as aggregate, and/or deicers applied in snow and ice removal.
Another recommendation set forth by the SHRP report is the use a non-reactive aggregate. This, however, is not always possible. There are limited aggregates with no potential for reactivity, since all silica has some potential for reaction and most aggregates contain significant amounts of silica. Recycled concrete when used as aggregate can also be reactive, particularly if it had already had ASR occurring before it was recycled. There are environmental reasons to use recycled glass as aggregate, but this is very reactive material. Also, transporting aggregates over long distances instead of using locally available materials adds significantly to the cost of concrete production.
Another recommendation is the use of appropriate levels of a suitable pozzolan. A pozzolan is a siliceous material that can combine with lime and water to harden, similar to a cement with just water. Since the hydration of cement produces lime as a byproduct (resulting in its basic nature), pozzolans can work well with cements. The pozzolan may be added as a mineral admixture at the time of concrete production, blended with the cement, or interground with the cement during the final production step of cement. The end result is about the same, as neither the cement nor the pozzolan is substantially changed as a result of the blending.
However, sources of suitable pozzolans are not always available locally, and the supplies are limited. Also, many pozzolans used for this purpose are waste products, and thus are quite variable in composition. An example is fly ash, which is the end result of coal burned for electric generation.
Further, sufficient amounts of the pozzolan must be used, or the protection is short lived, or the ASR can actually be made worse. This is particularly true of pozzolans with significant lime contents (&gt;10%), such as many fly ashes. In a cementitious system, the Ca:Si ratio is very important to its stability with regard to ASR. The higher the Ca:Si ratio, the less capable the system is of tying up alkali present. Thus the protection against ASR is less with high lime fly ashes. A low lime content pozzolan (&lt;10%) will reduce the ratio and give more protection from increased alkalies. However, a high lime content pozzolan will not give this protection, and further, since pozzolans carry their own alkalies into the system, this can easily make the situation worse.
Still another recommendation is the use a lithium-based admixture. Use of lithium was shown to be effective in ASR inhibition in 1951 (see W. J. McCoy and A. G. Caldwell, "New approach to inhibiting alkali-aggregate expansion," J.Amer.Concrete Institute, 22:693-706 (1951)). See also Y. Sakaguchi, et al., "The inhibiting effect of lithium compounds on alkali-silica reaction," Proceedings, 8th international conference, alkali aggregate reaction, Kyoto, Japan: 229-234 (1989), and the SHRP report.
Lithium salts, such as lithium hydroxide monohydrate, have been added to cement at the grinding stage of the cement production. J. Gajda, Development of a cement to inhibit alkali-silica reactivity, Skokie, Ill., Portland Cement Association Research and Development Bulletin RD115T (1996). As with the pozzolan blended cements mentioned above, the net effect is basically the same as if the lithium salt were admixed into the concrete separately at the time the concrete was batched. That is, by adding the lithium salt to the cement at the time of grinding, neither the cement nor the lithium salt is changed during the process. Rather, the lithium salt and cement are merely blended.
One problem associated with adding lithium salts to the concrete mixture or grinding lithium salts with cement is that substantial amounts of the lithium (about 50%, as reported in the SHRP Report) are "locked up" in the hydration products of the early hydration reactions of the cement. A substantial portion of the lithium ions, therefore, is unavailable for use in controlling ASR. This generally happens within the first few days of hydration. Still further, the concentration of hydroxyl ion can increase when lithium salts are admixed into concrete and mortars (see the SHRP Report). This can make the job of the lithium ion that much more difficult, and requires more lithium than would otherwise be the case.